Professional Documents
Culture Documents
Air Travel Demand: Iata Economics Briefing N 9
Air Travel Demand: Iata Economics Briefing N 9
Demand
IATA ECONOMICS BRIEFING No 9
Measuring the responsiveness
of air travel demand to changes
in prices and incomes
IATA Economics Briefing No 9:
AIR TRAVEL DEMAND
Mark Smyth
Brian Pearce
01 Executive Summary
page 04
02 Report Outline
page 12
07 Policy Implications
page 32
08 Conclusions
page 36
Summary
The demand for air travel is sensitive to
changes in air travel prices and incomes.
However, the degree of sensitivity (i.e.
its demand elasticity) will vary according
to different situations. To ensure that air
transport policies are effective, reliable
estimates for demand elasticities are essential.
This report provides important new evidence on the size and appropriate use for
demand elasticities for air travel. It outlines the key findings from extensive research
undertaken on behalf of IATA by InterVISTAS Consulting Inc. to estimate air travel
demand elasticities applicable to a wide range of air transport markets1. It builds upon
previous academic research to provide new estimates of air travel demand elasticities
in different locations and scenarios, based on a comprehensive econometric analysis
using key air travel databases.
The aim of the study is to provide robust air travel demand elasticity estimates to ensure
that policy decisions related to issues such as liberalisation, airport charges, taxation,
emissions schemes, are made on the basis of appropriate and reliable evidence. It provides
important new estimates to ensure that price elasticity estimates do not underplay the
sensitivity of passengers to price and are used correctly.
This report estimates the demand elasticity of air travel under various scenarios and
locations. It focuses on three main types of demand elasticity:
1
The full report “Intervistas (2008), Estimating Air Travel Demand Elasticities” contains more details on the
econometric results and can be found at www.iata.org/economics.
01 - Air Travel Demand 5
• Own price elasticity is a measure used to capture the There appears to be an inconsistency between the
sensitivity of consumers demand for a good or service size of price elasticities estimated for the air transport
in response to changes in the price of that particular industry and those estimated for the overall travel and
good or service. Goods with elasticities less than one tourism industry. But there are two main explanations
in absolute value are inelastic or price insensitive. for this. Firstly, as the air travel component of the
Goods with elasticities greater than one in absolute journey can be relatively easily substituted between
value are elastic or price sensitive. airlines, routes, modes, etc, the price elasticity for the
air travel price can be much higher than suggested by
• Cross price elasticity measures the interaction or the the price elasticity of the overall journey cost. Secondly,
sensitivity of demand for a particular good to changes passengers (especially for short leisure journeys) can
in the price of another good. When the cross price use a “two-stage” decision-making process, selecting a
elasticity is positive, the two goods are substitutes, flight destination based on the level of air travel price
when it is negative the goods are complementary. offered and then considering the other costs associated
with the journey.
• Income elasticity measures the sensitivity of demand
for a good to changes in individual or aggregate The appropriate value of a demand elasticity will vary in
income levels. accordance to the context in which they are considered.
For air transport there are five main levels (for the scope
of the market) for which demand elasticities can be
Air Travel Demand Elasticities estimated:
The elasticity of air travel demand varies according to the • Price Class Level. This the most disaggregate level,
coverage and location of the market in which prices are where passengers make a choice between different
changed and the importance of the air travel price within price classes (e.g. first class, business class, economy
the overall cost of travel. The appropriate elasticity to use class) on individual airlines.
will depend on the type of question being asked. What is
the price that is being changed (e.g. an individual airline • Airline / Air Carrier Level. This reflects the overall
ticket price or prices within the market as a whole)? What demand curve facing each airline on a particular route.
is the unit of demand that is being assessed (e.g. demand
• Route / Market Level. At the route or market level
for an individual airline or demand for total air travel)?
(e.g. London Heathrow–Paris CDG or London–Paris),
Examining the traffic impact of a price increase on a given
travellers faced with a price increase on all carriers
route requires a different elasticity than when examining
serving a route (e.g. due to an increase in airport fees
the impact of an across-the-board price increase on all
and charges), and have fewer options for substitution.
routes in a country or region.
• National Level. At the national level, travel prices are
There often appears to be some confusion in policy
increased on all routes to and from a particular country
discussions about the sensitivity of airline passengers to
(e.g. due to a higher national departure tax), giving
the price of travel. This has increased as the industry has
travellers fewer options for avoiding the price increase.
changed, with the Internet increasing price transparency,
deregulated markets and no frill carriers increasing • Supra-National Level. This represents a change in
competition and corporate travel buyers becoming more travel prices that occurs at a regional level across
price sensitive. In particular, there is an apparent paradox several countries (e.g. an aviation tax imposed on all
whereby: member states of the European Union). In this case,
the options for avoiding the price increase are even
• Passengers are becoming increasingly sensitive
further reduced.
to price, led by the boom in low cost travel, the
transparency brought by the Internet and the intense In each of the five levels of aggregation, different
competition on deregulated markets. cross-price elasticities exist, reflecting the availability of
substitute options. The own price elasticity at one level of
• But, passengers are also becoming less sensitive to
aggregation can reflect both the own price and cross price
price, as increasingly lower air travel prices, in real
elasticities at other levels of aggregation. The interaction
terms mean that the air travel price itself becomes a
between these effects adds significant complexity to the
smaller and less important part of the total cost of a
analysis, requiring clarity on which own price and cross
typical journey.
price elasticities were measured and controlled for. For The studies including the income term all produced
example, an analysis of route-level elasticities which does positive income elasticities, with air travel increasing at
not control for route substitution effects may be more a higher rate than income growth.
appropriate for a national-level elasticity.
• Route Level: -1.4 (e.g. travellers can switch to rail or car in response
to air travel price increases). While the geographical
• National Level: -0.8 breakdowns outlined in the next section capture some
variation by length of haul, there is still considerable
• Supra-National Level: -0.6
variation within each market. In particular, very short-
Route Level. The review of previous research found route haul flights (approximately less than 1 hour flight time)
level elasticities ranging from -1.2 to -1.5. Regressions are subject to greater competition from other modes.
using the US DB1A data, which allows the use of route
• On this basis an elasticity multiplier of 1.1 is used
dummies and variables to capture the price of route
to adjust air travel price elasticities for short-haul
substitutes, produced a similar air travel price elasticity
markets. This does not apply to the analysis of trans-
of -1.4. This elasticity estimate is applicable to a situation
Atlantic or trans-Pacific markets, which are entirely
where the price of an individual route changes (e.g.
long-haul, with virtually no opportunity for modal
higher airport charges at Paris CDG raising the price of
substitution.
travel from London and diverting leisure traffic to another
destination, such as Frankfurt). Using distance as an
instrument variable within the 2SLS (Two Stage Least
iii) Geographic Market Analysis
Squares) statistical model produces results that further The econometric analysis of the IATA PaxIS Plus data
support this elasticity, though there still is some concern found statistically significant differences between
over the use of distance in this way due to its perceived different geographic air travel markets. The estimated
exogenous influence on demand. elasticity multipliers for each market, along with the
reasons for why it is needed, are:
National Level. Estimates of national elasticities using all
three datasets found that, without the route substitution • Intra North America. This is our reference point
term, elasticities fell to around -0.8. This inelastic result with an elasticity multiplier of 1. The market is well
was found over a range of model specifications which established with relatively high levels of capacity and
excluded route dummies. The national level elasticity traffic. Prices tend to be low, while distances are short
applies to a situation such as the doubling of a national to medium haul.
passenger departure tax, affecting all departing routes
equally but leaving the cost of travel from elsewhere • Intra Europe. Traffic in this region is estimated to be
unchanged. Its value reflects a combination of the route more elastic, with an elasticity multiplier of 1.4. Intra
own price elasticities with cross price elasticities, when European routes typically have shorter average travel
all national routes have prices which vary in the same distances, strong competition from other transport
way. The inelastic result is consistent with observations modes and the use of very low prices in several
that part of the price elasticity observed from low cost markets. The high market share of charter airlines is
carriers (LCCs) involves substitution from other routes. being eroded by very low fare LCCs.
When this is controlled for, LCCs have a lower level of
• Intra Asia. Moderately more inelastic estimates were
market stimulation, consistent with less elastic national
found in this region, with an elasticity multiplier of
elasticities.
0.95. LCCs are now emerging in Asia but average
Supra-National Level. At the supra-national level distances are longer, and the key middle class is still
(e.g. the European Union) estimates show an even less relatively small in many markets in this region.
elastic air travel price elasticity of -0.6. This is because
• Intra Sub-Saharan Africa. This region is estimated to
as the number of routes covered expands the number of
have a relatively inelastic demand compared to North
choices for passengers to avoid any travel price increase
America, with an elasticity multiplier of 0.6. African
diminishes. There is less opportunity for traffic to be
economies have a much smaller middle class. Travel
diverted.
is concentrated among higher income individuals who
are less price-sensitive.
ii) Short-Haul vs. Long-Haul
• The review of previous research found consistent • Intra South America. This region is estimated to be
results showing that air travel price elasticities on at the more elastic end of the scale, with an elasticity
short-haul routes were higher than on long-haul multiplier of 1.25. There is an emerging middle class
routes. This largely reflects the greater opportunity making the region more price elastic plus LCCs are
for inter-modal substitution on short haul routes emerging in Brazil, Chile and Mexico.
• Trans Atlantic (North America – Europe). A high By way of illustration, elasticities for different situations
price elasticity was found for this market, with an can be developed by selecting the relevant base price
elasticity multiplier of 1.2. This market has long been elasticity and applying the relevant multipliers. For
developed by low fare charter airlines. Price is likely example:
more important than frequency in this market than in
US domestic markets • To examine the impact of an EU-wide aviation tax on
short-haul markets, the elasticity would be developed
• Trans Pacific (North America – Asia). By contrast, as follows:
markets across the Pacific are estimated to have a
much less elastic response, with an elasticity multiplier - Base multiplier: -0.6 (supra-national)
- Geographic market: 1.4 (Intra Europe)
of 0.6. There are no charter services and there remain
- Short-haul adjustor: 1.1
markets with less liberal pricing regulation. There are The price elasticity would then be calculated as:
early signs of long-haul LCCs emerging but at present -0.6 x 1.4 x 1.1 = -0.92
this market shows much less sensitivity to travel price
than the US domestic market or the trans Atlantic • To examine the impact of a UK tax on aviation
market. on Trans Atlantic traffic, the elasticity should be
developed as follows:
• Europe-Asia. This market is estimated to be slightly
less price sensitive, with an elasticity multiplier of - Base multiplier: -0.8 (national)
- Geographic market: 1.2 (Trans Atlantic)
0.9. This result is in contrast to the results found in
The price elasticity would then be calculated as:
the respective intra markets of Europe and Asia, and -0.8 x 1.2 = -0.96
provides further evidence for lower elasticities on long-
haul and intercontinental air transportation. • To examine the impact of an increase in airport landing
fees on a particular short-haul route in Asia, the
elasticity should be developed as follows:
Applying the Elasticity Estimates - Base multiplier: -1.4 (route)
and Multipliers - Geographic market: 0.95 (Intra Asia)
- Short-haul adjustor: 1.1
Table 1 provides a guideline for the estimated price The price elasticity would then be calculated as:
demand elasticity by level of aggregation and by -1.4 x 0.95 x 1.1 = -1.46
region. It multiplies the estimate for the relevant level of
aggregation by the relevant short-haul and geographic
elasticity multipliers.
2
This conclusion was reached in a recent study of the UK market (UK CAA, 2006, No frills carriers; revolution or evolution). It concludes that
“despite the spectacular growth of no-frills carriers in the UK, and the perceptions about the impact they have had on travel habits, there has
been little change in long-term aggregate passenger traffic growth rates”.
3
Building a Greener Future (2007) IATA.
01 - Air Travel Demand 11
Policy Implications The focus of existing policy to reduce CO2 emissions from
air travel has been on trying to manage air travel demand
The correct elasticity value to use in analysing an air by raising the cost of travel for passengers. The results
transport policy decision depends on the type of question contained in this report show that such policies are likely
being asked. The impact on demand of higher travel to fail. Decoupling emissions from travel growth needs to
costs on a given route due to a rise in airport landing focus not on demand management but on mechanisms
charges requires a different (higher) elasticity than when to bring about emissions reduction measures from
examining the traffic impact of a wider travel cost increase technology, infrastructure and operations.
due to a passenger tax on all routes in a country.
IATA’s 4-Pillar Climate Strategy3 , which was endorsed
Air transport policy decisions run the risk of being by the Assembly of the International Civil Aviation
ineffective, or even counter-productive, if the correct Organisation in 2007, focuses action on emission
demand elasticity is not used. For example: reduction measures from technology, infrastructure,
operations and those brought about by well designed
• A revenue-raising policy to raise the price of travel economic instruments.
on a route (e.g. higher airport charges) will reduce
passenger numbers more than expected if the price Effectively decoupling emissions from air travel growth
elasticity of demand is underestimated. A price elastic will require policy-makers and the industry to look beyond
response to air travel price increases at the route level simple economic instruments:
means that demand falls at a proportionately higher
• Technological progress will require collaboration
rate than the increase in price.
across the value chain and across countries.
• A national passenger tax may damage inbound
• Governments will need to play a role in funding
tourism more than expected if policy-makers do
fundamental research.
not take into account the greater price elasticity of
overseas residents visiting the country, who have • Political will is perhaps one of the most important
a choice of destinations, compared to outbound mechanisms for delivering emissions reductions from
domestic residents who can either pay the tax or infrastructure improvements.
not travel.
• The lack of implementation of a Single European
• An environmental policy to raise the price of travel Sky is one glaring omission in policy action to reduce
(e.g. a national aviation tax) on a national or supra- emissions from air travel.
national basis will be ineffective if the price elasticity
of outbound travel is low, as found. A price inelastic IATA is actively promoting collaborative efforts on
response at the national or supra-national level technology and is lobbying hard for governments to
means that the impact in terms of reducing demand improve infrastructure. For operations, there is a major
will be proportionately less than the increase in price. initiative to spread best practices. More needs to be done
The greater price sensitivity of inbound overseas about the challenges of climate change, but the airline
residents will result in a diversion to other destinations, industry is already stepping up efforts with a bold vision of
a “leakage of carbon”, and thus a reduction in zero emissions and an important target of carbon-neutral
environmental effectiveness. growth. The key lesson for both policy-makers and the
industry is to look beyond simple economic instruments
for mechanisms to bring about an effective reduction in
emissions from air travel.
02 Report Outline
This report discusses the key findings from extensive Previous estimates
research undertaken on behalf of IATA by InterVISTAS
Consulting Inc. to estimate air travel demand elasticities Chapter 5 summarises the key findings from the review
applicable to a wide range of air transport markets. undertaken by InterVISTAS Consulting Inc. of previous
It provides important new evidence for estimating academic studies that have estimated air travel demand
elasticities in different locations and scenarios, based on elasticities. It outlines the wide range of estimates that
access to comprehensive air travel databases. It builds have been produced, as well as the common themes
upon previous academic research into air travel demand between the majority of the studies.
elasticities.
New evidence for air travel demand elasticities
The report is structured as follows:
Chapter 6 outlines the key findings from a comprehensive
econometric analysis of the data from three key air travel
What are Demand Elasticities?
data sources. The results from this analysis provide
Chapter 3 provides definitions for different categories important new evidence of the relevant demand elasticity
of demand elasticities and discusses how each can in different markets.
be interpreted. It outlines three key concepts: price
elasticity of demand, cross-price elasticity of demand What does this mean for air transport policymakers?
and income elasticity of demand. It also discusses key
Chapter 7 discusses how this new evidence can be applied
factors that influence both the demand for air transport
to air transport policy decisions. It provides guidelines
and the sensitivity of this demand to changes in prices
for the application of air travel demand elasticities for
and incomes.
different types of decisions. It helps to ensure that policy
decisions are made on the basis of appropriate and
How can this be applied to air travel demand?
reliable estimates of their impact on demand.
Chapter 4 discusses the different types of demand
elasticity that can be estimated for air travel. It discusses Chapter 8 provides a summary and conclusions.
how the relevant demand elasticity can change in relation
to the market size and its geographical location.
03 The Definition
of Demand
Elasticities
Elasticity measures the response of one
economic variable (e.g. the quantity
demanded) to a change in another economic
variable (e.g. price). It provides key insight
into the proportional impact of different
economic actions and policy decisions.
03 - Air Travel Demand 15
Demand elasticities measure the change in the quantity ii) Cross Price Elasticity of Demand
demanded of a particular good or service as a result of
changes to other economic variables, such as its own The cross price elasticity measures the interaction in
price, the price of competing or complementary goods and demand between different goods or services. It measures
services, income levels, taxes, etc. This report estimates the sensitivity of demand for a particular good to changes
the demand elasticity of air travel under various scenarios in the price of another good:
and locations. It highlights three main types of demand Cross Price Elasticity =
elasticity: own price elasticity, cross price elasticity and
income elasticity. % Change in Quantity of Good A Demanded
% Change in Price of Good B
i) Price Elasticity of Demand When the cross price elasticity is positive, the two goods
Price elasticity is a measure used to capture the sensitivity are substitutes (e.g. Coca-Cola and Pepsi). In other words,
of consumer demand for a good or service in response to an increase in the price of one good will lead consumers
changes in the price of that particular good or service. to shift demand towards the relatively cheaper substitute
The price elasticity is defined as: good. When the cross price elasticity is negative the
goods are complementary goods (e.g. coffee and milk).
Price Elasticity = In other words, an increase in the price of one good will
% Change in Quantity Demanded negatively affect both its own demand and the demand of
% Change in Price goods that are usually bought to accompany it.
The quantity demanded generally decreases when the iii) Income Elasticity of Demand
price increases, so this ratio is usually expected to be
negative. For example, if a 10% increase in the price of The income elasticity measures the sensitivity of demand
good A results in a 6% fall in the quantity demanded of for a good to changes in individual or aggregate income
that good, its own price elasticity is -0.6. By contrast, if a levels:
10% fall in the price of good B leads to a 12% increase in
Income Elasticity =
the quantity demanded of good B, its own price elasticity
is -1.2. % Change in Quantity Demanded
% Change in Income
Goods with elasticities less than one in absolute value
are commonly referred to as having inelastic or price An income elasticity between 0 and +1 indicates a
insensitive demand. In other words, the proportional normal good, where the quantity demanded increases at
change in quantity demanded will be less than the the same or a lesser rate than the increase in income. For
proportional change in price. In this situation, increasing example, a good where a 10% increase in income results
the price will increase the revenue received by the in a 0-10% increase in consumption would be considered
producer of the good, since the revenue lost by the a “normal” good.
decrease in quantity is less than the revenue gained from
An income elasticity greater than +1 indicates what
the higher price.
economists call a luxury good, where consumption
Goods with elasticities greater than one in absolute increases by a greater proportion than income. For
value are referred to as having elastic or price sensitive example, as discretionary incomes rise consumers can
demand. In other words, the proportional change in afford to buy higher quality and/or leisure related goods
quantity demanded will be greater than the proportional that were previously beyond their reach. This does not
change in price. A price increase will result in a revenue mean these goods are the exclusive preserve of the rich,
decrease to the producer since the revenue lost from but that as living standards rise consumers value buying
the resulting decrease in quantity sold is more than the these goods the most. It is a measure of a highly valued
revenue gained from the price increase. good in consumer welfare terms.
A negative income elasticity indicates an inferior good,
where the quantity demanded decreases as aggregate
incomes increase. In other words, with higher incomes,
consumers buy less of an inferior good and substitute
it with better quality goods (e.g. buying branded goods
rather than supermarket own-brands).
Travel Demand
Elasticities
The elasticity of air travel demand varies
according to the coverage and location of
the market in which prices are changed
and the importance of the air travel price
within the overall cost of travel. The
appropriate elasticity to use will change
according to the precise impact that is
being analysed.
The previous chapter provided definitions for the different types of demand elasticity.
Yet, when using this methodology to estimate the demand elasticity of air travel it is
important to be clear about the precise question that is being asked. What is the price
that is being charged (e.g. an individual airline ticket price or prices within the market
as a whole)? What is the unit of demand that is being assessed (e.g. demand for an
individual airline or demand for total air travel)? Estimating an appropriate demand
elasticity requires a clear definition of what is being studied.
04 - Air Travel Demand 19
4
For example, the International Passenger Survey by the UK Office of National Statistics estimates that for UK residents travelling to Europe
by air, the travel price represented 27% of the total journey cost in 2004. This proportion has declined in the last ten years, led by the growth of
no-frills LCC airlines.
For air transport there are five main levels (for the scope The evidence and discussion provided in this report focus
of the market) for which demand elasticities can be on the appropriate elasticities for the route, national and
estimated: supra-national level of aggregation.
• Price Class Level. This is the most disaggregate level, The Interaction between Own-Price
where passengers make a choice between different and Cross-Price Elasticites
price classes (e.g. first class, business class, economy
class) on individual airlines. At this level, the elasticities In each of the five levels of aggregation, different
are arguably highest, with passengers easily able to cross-price elasticities exist, reflecting the availability of
switch between price-class levels and airlines, while substitute options. For example:
also (in some cases) having the option to use another
• At the price class level, an increase in the full economy
mode of travel or simply to choose to not travel (i.e.
price could increase the demand for both business
other activities act as a substitute for air travel).
class tickets and discount tickets.
• Airline / Air Carrier Level. This reflects the overall
• At the airline level, a unilateral increase in the travel
demand curve facing each airline on a particular route.
price of one particular airline on a route can increase
Where there are a number of airlines operating on a
the demand for other carriers on the route (and the
route , the demand elasticity faced by each airline is
demand for connecting alternatives).
likely to be fairly high. If an airline increases its price
unilaterally, it is likely to lose passengers to other • At the route level, an increase in the price of travel
airlines operating on that route.5 from London Heathrow to Paris CDG can increase the
demand for travel on London Gatwick to Paris CDG or
• Route / Market Level. At the route or market level
London Heathrow to Paris Orly.
(e.g. London Heathrow–Paris CDG or London-Paris),
the elasticity response is expected to be lower than at • At the national level, an increase in the price of air
the price class or carrier level. Travellers faced with a travel to/from a given country may increase demand
travel price increase on all carriers serving a route (e.g. for air travel to/from other countries.
due to an increase in airport fees and charges), have
fewer options for substitution. However, they can still • At the supra-national level, an increase in the price
choose to travel on an alternative route, while also (in of air travel to/from a particular region may increase
some cases) having the option to use another mode of demand for air travel to/from other regions (e.g.
travel or simply choose to not travel. an increase in the cost of air travel to the EU may
increase demand for air travel to the US).
• National Level. At the national level, travel price
elasticities are expected to be lower, as travellers • At all levels of aggregation, there may exist cross
have fewer options for avoiding the price increase. For elasticity effects with other modes of transport. An
example, if a national government imposed a new or increase in the price of air travel may increase demand
increased tax on aviation, travellers could only avoid for ground transportation and vice versa.
this increase by travelling elsewhere, using another
• There may also be cross elasticity effects between
mode (which may not always be possible), or choosing
air travel and other leisure or consumption activities.
not to travel. For example, if the UK government
In some cases it may not exist at all (e.g. there is
imposed an increased tax on aviation departures, UK
generally no substitute for air travel on long-haul
residents travelling to mainland Europe could respond
routes).
by travelling by Eurostar or by ferry, or choose not to
travel. Similarly, travellers in France could respond by The own price elasticity at one level of aggregation can
travelling to the UK by another mode or by switching reflect both the own price and cross price elasticities
their destination to another country, such as Germany at other levels of aggregation. For example, the price
or Spain. elasticity at the route level is a function of the own
price and cross price elasticities at the price class
• Supra-National Level. This represents a change in
and carrier levels of aggregation6 . The interaction
prices that occurs at a regional level across several
between these effects adds significant complexity to
countries. For example, an aviation tax imposed on
the analysis, requiring clarity on which own price and
all member states of the European Union. In this
cross price elasticity were measured and controlled for.
case, the elasticity is expected to be even lower, as
For example, an analysis of route-level elasticities which
the options for avoiding the price increase are even
does not control for route substitution effects may be
further reduced.
more appropriate for a national-level elasticity.
5
Even in a situation where an air carrier has a monopoly on a route, it may still face a fairly high demand elasticity as connecting options can
also act as a substitute (e.g. passengers could travel from London to Delhi via Paris or Dubai rather than direct).
6
See Annex A for more details.
05 Previous
Estimates
of Air Travel
Demand
Elasticities
The results from previous studies of air travel
demand elasticities vary in accordance
with the type of elasticity that is estimated.
Nevertheless, there are several key themes
that can be identified that provide insights
into the sensitivity of air travel demand.
05 - Air Travel Demand 23
There is a substantial amount of evidence on air travel • Business Versus Leisure Passengers. In general,
demand elasticities available from previous studies. all else being equal, business travellers are less
However, these studies have also produced a wide variety sensitive to travel price changes (less elastic) than
of results, reflecting large differences in the nature and leisure travellers. Intuitively, this result is plausible;
scope of the elasticities that have been looked at. As business travellers generally have less flexibility to
part of its research, InterVISTAS Consulting reviewed the postpone or cancel their travel than leisure travellers.
available literature on demand elasticities. This chapter Nevertheless, the studies do show that even business
summarises the key themes that can be identified across travel will decline in the face of price increases, albeit
the different studies. not to the same extent as leisure travel.
7
See Annex B for a list of the studies that were reviewed.
8
Gillen, Morrison and Stewart (2002), Air Travel Demand Elasticities: Concepts, Issues and Measurements.
06 New Estimates
of Price
and Income
Elasticities
of Air Travel
Demand
New econometric research builds on the
work of previous studies to show that
demand elasticities vary according to the
level of aggregation and the location of
the market. It develops guideline demand
elasticity estimates to use for different
situations.
06 - Air Travel Demand 25
The literature review highlighted that air travel price Model Specifications
elasticities vary depending on a number of factors such
as location, distance and level of market aggregation. The new econometric research tested a number of
InterVISTAS Consulting used this guidance to undertake different estimation methods and model specifications
new econometric analysis using three different datasets. to provide explanations of economic phenomena within
The new analysis is comprehensive, based on over 500 each given market. In all models, traffic was the dependent
regression models. Building on the previous research, variable. The different specifications used were9:
the new econometric analysis develops a set of in-depth • Ordinary Least Squares (OLS). OLS is a method
guidelines, and guideline elasticities, that can be applied relating passenger traffic to air travel prices,
to the analysis of different air markets, broken down by income levels and other variables, while minimising
geographic market, length of haul and level of aggregation the variance (randomness) of the estimates. The
(route, national and supra-national). regression analysis allows the relationship between
traffic and air travel prices to be isolated and
Data Sources quantified while controlling for other factors that may
InterVISTAS Consulting used three different databases impact air travel, such as GDP, population levels, route
for the econometric research: distance and seasonality.
9
See Annex C for more details on each type of specification.
converted into US dollars. GDP appears in all model whether there is a low cost carrier, what types of
specifications. cultural or financial linkages there are between pairs
of cities). The use of route dummy variables controls
• Population. Population has a direct effect on the size for these variables without the need to quantify them.
of a market and may cause a bias in the estimates
if omitted. For example, a large increase in traffic
may reflect a sudden boom in population rather than
other effects. Population was tested in all model
ECONOMETRIC RESULTS
specifications but the best results tended to be at the The new econometric research supports the discussion
city pair level, so it is only included in the US domestic from previous chapters that the sensitivity of passengers
regression results. to the level of air travel prices depends significantly on the
level of the market being considered and its location. The
• Route Distance (Trip Length). The use of route
results are based on a synthesis of the new econometric
distance is based on its ability to reflect the value of
results and the review of previous research. Base
travel time savings and availability of substitutes. As
elasticity estimates are developed for the different levels
distance increases, the viability of other transport
of aggregation (route, national and supra-national level).
modes as a substitute decreases. The use of route
Multiplicative estimates were then developed to adjust
distance as an instrumental variable in 2SLS requires
the elasticities to reflect specific geographical markets.
that distance be uncorrelated with traffic. This is most
likely in the domestic US market. i) Level of Aggregation
• Substitute Goods. The inclusion of a substitute travel In summary, the econometric results found that at the
price variable was tested on a subset of routes and route level (where competition between airlines or city-
was found to increase price elasticities. Estimates pair markets is high) the sensitivity of demand to price
that exclude a meaningful substitute in the regression is very high. However, at the national or regional level,
model will produce more inelastic estimates than air travel is relatively price insensitive. The results support
a correctly specified model would produce. Route demand elasticities of:
substitutes can be defined as a different airport
serving the same catchment area (e.g. Chicago- • Route Level: -1.4
O’Hare or Chicago-Midway) or a different destination
• National Level: -0.8
serving the same purpose (e.g. Las Vegas or Reno).
• Supra-National Level: -0.6
• Real Exchange Rates.10 In theory, as the foreign
country becomes more expensive (inexpensive), Route Level. The review of previous research found route
leisure travellers will travel to the foreign country level elasticities ranging from -1.2 to -1.5. Regressions
less (more). However, the econometric research using the US DB1A data, which allows the use of route
was unable to obtain robust estimates using this dummies and variables to capture the price of route
variable, possibly due to difficulties in obtaining an substitutes, produced a similar air travel price elasticity
accurate measure of the variable. Therefore, real of -1.4. This elasticity estimate is applicable to a situation
exchange rates were excluded from the final model where the price of an individual route changes (e.g.
specifications. higher airport charges at Paris CDG raising the price of
travel from London and diverting leisure traffic to another
• Time Variables. The use of several different forms
destination, such as Frankfurt). Using distance as an
of time variables was explored, but only quarterly
instrument variable within the 2SLS model produces
(seasonal) time dummy variables were found to
results that further support this elasticity, though there
increase the explanatory power of the model. This
still is some concern over the use of distance in this way
suggests that both travel prices and demand are
due to its perceived exogenous influence on demand.
inherently seasonal (at least on a quarterly basis).
National Level. Estimates of national elasticities using all
• Route Dummies. These dummy variables were found
three datasets found that, without the route substitution
in many cases to increase the explanatory power of
term, elasticities fell to around -0.8. This inelastic result
the model, suggesting that some other route-specific
was found over a range of model specifications which
factors are important for demand (e.g. whether or
excluded route dummies. The national level elasticity
not non-stop service is available, whether there are
applies to a situation such as the doubling of a national
competing carriers operating on non-stop routes,
10
The real exchange rate is defined as the consumer price index (CPI) in the foreign country divided by the CPI in the domestic country divided
by the exchange rate (foreign currency per unit of domestic currency).
06 - Air Travel Demand 27
passenger departure tax, affecting all departing routes • Intra Europe. Traffic in this region is estimated to be
equally but leaving the price of travel from elsewhere more elastic, with an elasticity multiplier of 1.4. Intra
unchanged. Its value reflects a combination of the route’s European routes typically have shorter average travel
own price elasticities with cross price elasticities, when distances, strong competition from other transport
all national routes have prices which vary in the same modes and the use of very low prices in several
way. The inelastic result is consistent with observations markets. The high market share of charter airlines is
that part of the price elasticity observed from low cost being eroded by very low fare LCCs.
carriers (LCCs) involves substitution from other routes.
When this is controlled for, LCCs have a lower level of • Intra Asia. Moderately more inelastic estimates were
market stimulation, consistent with less elastic national found in this region, with an elasticity multiplier of
elasticities. 0.95. LCCs are now emerging in Asia but average
distances are longer, and the key middle class is still
Supra-National Level. At the supra-national level relatively small in many markets in this region.
(e.g. the European Union) estimates show an even less
elastic air travel price elasticity of -0.6. This is because • Intra Sub-Saharan Africa. This region is estimated to
as the number of routes covered expands, the number of have a relatively inelastic demand compared to North
choices for passengers to avoid any travel price increase America, with an elasticity multiplier of 0.6. African
diminishes. There is less opportunity for traffic to be economies have a much smaller middle class. Travel
diverted. is concentrated among higher income individuals who
are less price-sensitive.
ii) Short-Haul Vs Long-Haul • Intra South America. This region is estimated to be
The review of previous research found consistent results at the more elastic end of the scale, with an elasticity
showing that air travel price elasticities on short-haul multiplier of 1.25. There is an emerging middle class
routes were higher than on long-haul routes. This largely making the region more price elastic plus LCCs are
reflects the greater opportunity for inter-modal substitution emerging in Brazil, Chile and Mexico.
on short-haul routes (e.g. travellers can switch to rail or
car in response to air travel price increases). While the • Trans Atlantic (North America – Europe). A high
geographical breakdowns outlined in the next section price elasticity was found for this market, with an
capture some variation by length of haul, there is still elasticity multiplier of 1.2. This market has long been
considerable variation within each market. In particular, developed by low fare charter airlines. Price is likely
very short-haul flights (approximately less than 1 hour more important than frequency in this market than in
flight time) are subject to greater competition from other US domestic markets
modes.
• Trans Pacific (North America – Asia). By contrast,
On this basis an elasticity multiplier of 1.1 is estimated markets across the Pacific are estimated to have a
and used to adjust air travel price elasticities for short- much less elastic response, with an elasticity multiplier
haul markets. This does not apply to the analysis of trans of 0.6. There are no charter services and there remain
Atlantic or trans Pacific markets, which are entirely long- markets with less liberal pricing regulation. There are
haul, with virtually no opportunity for modal substitution. early signs of long-haul LCCs emerging but at present
this market shows much less sensitivity to travel cost
iii) Geographic Market Analysis than the US domestic market or the trans Atlantic
market.
The econometric analysis of the IATA PaxIS Plus data
found statistically significant differences between • Europe – Asia. This market is estimated to be slightly
different geographic air travel markets. The estimated less price sensitive, with an elasticity multiplier of
elasticity multipliers for each market, along with the 0.9. This result is in contrast to the results found in
reasons for why it is needed, are: the respective intra markets of Europe and Asia, and
provides further evidence for lower elasticities on long-
• Intra North America. This is our reference point haul and intercontinental air transportation.
with an elasticity multiplier of 1. The market is well
established with relatively high levels of capacity and
traffic. Air travel prices tend to be low, while distances
are short to medium haul.
Applying the Elasticity Estimates and Multipliers
Table 3 provides a guideline for the estimated price demand elasticity by level of aggregation and by region. It
multiplies the estimate for the relevant level of aggregation by the relevant short-haul and geographic elasticity
multipliers.
Njegovan (2006) found that UK residents travelling • To examine the impact of an increase in airport landing
overseas for leisure had a price elasticity of -0.7 to -0.8. fees on a particular short-haul route in Asia, the
Business travel will be less elastic. So a conservative elasticity should be developed as follows:
estimate for the price elasticity (Pe) of inbound travel by
overseas residents can be derived as follows: - Base multiplier: -1.4 (route)
- Geographic market: 0.95 (Intra Asia)
Peoverseas residents * % overseas residents +
PeUK residents * % UK residents = average Pe - Short-haul adjustor: 1.1
Peoverseas residents * 39% + -0.8 * 61% = -1.0 The price elasticity would then be calculated as:
-1.4 x 0.95 x 1.1 = -1.46.
Peoverseas residents = -1.3
11
This conclusion is reached in a recent study of the UK market (UK CAA, 2006, No frills carriers; revolution or evolution). It concludes that “despite
the spectacular growth of no-frills carriers in the UK, and the perceptions about the impact they have had on travel habits, there has been little
change in long-term aggregate passenger traffic growth rates”.
07 Policy
Implications
Air transport policy decisions run the risk
of being ineffective, or even counter-
productive, if the correct demand elasticity
is not used. A revenue-raising policy to
raise the cost of travel on a route will
be ineffective if the demand elasticity is
underestimated. An environmental policy
to raise the price of travel on a national or
supra-national basis will be ineffective if
the demand elasticity is overestimated.
07 - Air Travel Demand 33
Base elasticity -0.6 (supra-national) multiplied by IATA’s four pillar climate strategy12, which was endorsed
1.4 (intra-Europe geographic multiplier) multiplied by the Assembly of the International Civil Aviation
by 1.1 (short-haul multiplier) which equals -0.92 Organisation in 2007, focuses action on emission
reduction measures from technology, infrastructure,
So a 10% rise in intra-EU travel costs would lead to a
operations and those brought about by well designed
relative inelastic 9.2% reduction in air travel.
economic instruments.
12
Building a Greener Future (2007) IATA.
By contrast, Table 5 shows that: Effectively decoupling emissions from air travel growth
will require policy-makers and the industry to look beyond
• Passenger taxes are ineffective for reducing CO2
simple economic instruments:
emissions. This is not just because demand is relatively
price insensitive at a national and supra-national • Technology progress will require collaboration across
level. It is also because raising the cost of travel the value chain and across countries.
for the passenger does nothing to incentivise the
manufacturer to produce new airframes or engines, • Governments will need to play a role in funding
nothing to incentivise the fuel company to produce a fundamental research.
clean fuel, nothing to incentivise the EU to implement
• Political will is perhaps one of the most important
a Single European Sky, nothing to incentivise ANSPs
mechanisms for delivering emissions reductions from
to straighten routes and reduce stacking, nothing to
infrastructure improvements.
incentivise airports to reduce taxiing emissions, and
nothing to incentivise airlines to improve operations • The lack of implementation of a Single European
and renew their fleet. Sky is one glaring omission in policy action to reduce
emissions from air travel.
• Emissions trading can be more effective, if well
designed. By a direct linkage to emissions it IATA is actively promoting collaborative efforts on
incentivises operational and fleet improvements. technology and is lobbying hard for Governments to
If well designed, to be open to trading with other improve infrastructure. On operations there is a major
industries and global, it allows the reduction of CO2 initiative to spread best practice. More needs to be done
emissions to take place in industries where reductions in the face of the challenge of climate change, but the
are most efficient. However, even emissions airline industry is already stepping up its efforts with a
trading has little impact on the key technology and bold vision of zero emissions and an important future
infrastructure pillars. milestone of carbon-neutral growth. The key lesson for
both policy-makers and the industry is to look beyond
simple economic instruments for mechanisms to bring
about an effective reduction in emissions from air travel.
This report uses new econometric research and a Thus, the particular elasticity value to be used for
review of previous studies to provide guidelines on the analysing price effects in airline markets depends
appropriate level and type of demand elasticity to use on the question being asked. The narrower the
when analysing a policy proposal. It provides robust applicability of a price change, the more elastic (i.e.
elasticity estimates to ensure that policy decisions larger) the change in demand. The more general the
related to issues such as liberalisation, airport charges, applicability of a price change (perhaps due to higher
taxation, and emissions schemes, are made on the costs or taxes) the less elastic (i.e. smaller) the change
basis of appropriate and reliable evidence. It provides in demand.
important new estimates to ensure that price elasticity
estimates do not underestimate the sensitivity of
passengers to price and are used correctly.
Different air travel demand elasticities are associated
with different uses. When consumers are choosing
between airlines on a route, or even between
destinations for travel, there is a degree of price
elasticity for airline seats. However, if all competitors
on a route, or if a wide range of routes all experience
the same proportionate price increase, the demand for
airline services becomes less elastic. As a price increase
is extended to ever larger groups of competing airlines
or competing destinations, then the overall demand for
air travel is revealed to be somewhat inelastic.
The implications are:
• For an airline on a given route, increasing price is
likely to result in a more than proportionate decrease
in air travel. Lower travel prices will greatly stimulate
traffic and raise revenues. Airline specific travel price
charges are price elastic.
Table A1: Impact of fare increases at the route and market level
Effect on route A Effect on route B
13
Demand for Outbound Leisure Air Travel and its Key Drivers (December 2005) UK CAA.
Air Travel Demand 39
ANNEX B: LIST OF PREVIOUS Oum, Zhang, and Zhang (1993), Inter-Firm Rivalry and
Firm-Specific Price Elasticities in Deregulated Airline
STUDIES REVIEWED Markets.
Abed, Ba-Fail and Jasimuddin (2001), An Econometric PriceWaterhouseCoopers (2005), Aviation Emissions
Analysis of International Air Travel Demand in Saudi and Policy Instruments, Final Report.
Arabia.
Rubin and Joy (2005), Where are the Airlines Headed?
Abrahams (1983), A Service Quality Model of Air Travel Implications of Airline Industry Structure and Change for
Demand: An Empirical Study. Consumers.
Alperovich and Machnes (1994), The Role of Wealth in Taplin (1980), A Coherence Approach to Estimates of
the Demand for International Air Travel. Price Elasticities in the Vacation Travel Market.
Australian Bureau of Transport and Communications Taplin (1997), A Generalised Decomposition of Travel-
Economics (1995), Demand Elasticities for Air Travel to Related Demand Elasticities into Choice and Generation
and from Australia. Components.
14
Log-log model formulations refer to a model specification where both the dependent (left hand side) and independent (right hand side) vari-
ables have been transformed by the natural logarithm.
15
A consistent estimator in a regression analysis is defined as an estimator that converges in probability to the correct parameter as the
sample size grows.
Air Travel Demand 41
16
The U.S. domestic market is not constrained by air bilateral agreements, has a wide access to capital, and high levels of competition, all of
which increase responsiveness to changes in the explanatory variables (travel price, income, and population).
Understanding the impact
on demand is the key to
effective policy decisions
concerning aviation.
www.iata.org
PRINTED IN SWITZERLAND